Head Start is the largest federal program providing an enriched early childhood education for children from low income families. A substantial body of non-experimental and quasi-experimental research has linked Head Start participation with (often modest) gains in children's developmental outcomes. Yet research to date has failed to examine how variability across Head Start centers is associated with variability in children's developmental outcomes, and how the quality of home and school environments experienced after Head Start might sustain, or curtail, the impacts of Head Start over time. To address this knowledge gap, the proposed project goes beyond questions of simple impact to consider the conditions and contexts which make Head Start more or less effective. Specifically, we will examine the extent to which the structure and quality of Head Start centers, parenting behavior and the home environment, and the structure and quality of elementary schools might mediate or moderate program impacts over time. Our interdisciplinary team (from the fields of human development, education, economics, and social work) will utilize two large, national Head Start studies- -one of which used an experimental design--to address the following aims: Aim 1: To identify to what extent, and by what processes, aspects of Head Start quality promote children's cognitive development, social-emotional skills, and physical health; Aim 2: To determine the role of parents in creating and sustaining positive long-term impacts of Head Start on children's cognitive development, social-emotional skills, and physical health; and Aim 3: To examine the extent to which subsequent school experiences moderate the persistence of Head Start effects on children's cognitive development, social-emotional skills, and physical health. The project involves secondary data analysis of two large, federally sponsored data-sets, namely the Head Start Family and Child Experiences Survey, 1997 Cohort (FACES-97), and the Head Start Impact Study (HSIS). Each study included a nationally representative sample of 3- and 4-year-old low income children attending Head Start, along with one control group of children on waiting lists for Head Start in the HSIS. Research questions will be addressed using a combination of multiple regression, piecewise regression, latent class growth analysis, and multiple group analysis. Of particular interest will be interactions between treatment condition in the HSIS and center quality in the preschool year and school quality in the elementary school years.